Today’s report (week ended 4 Jan): EIA: net change: -91 bcf, implied flow: -87 bcf, EA: -69 bcf
- Expectations for today’s estimate spanned a wide range. To align with today’s figure, we reduced our estimate for the holiday dent in industrial demand and upwardly revised our estimate for power demand. Major power markets had generation from gas-fired units increase by more than 10 aGW w/w, with most of that in the first few days of 2019.
Next Thursday’s report (week ending 11 Jan): EA preliminary: -80 bcf
- Holiday-related impacts are no longer weighing on demand, while production looks to have remained essentially flat w/w. However, the w/w decline in GWHDDs will push res-com demand down, leading to a lower stockdraw on the order of 1 bcf/d w/w.
Swings and merry-go-rounds
At the start of the heating season, each Panorama edition seemed to coincide with another chip off end-March 2019 inventories due to early-season cold. With the start of the new year, the opposite trend appears to be taking hold, with forecast cold continually pushed further into the horizon. End-March storage is on track for near 1.55 tcf, based on the 15-day forecast and 10-year normal weather, for the reminder of Q1 19. If a carryout near 1.55 tcf is realised, the current 2019 injection strip and Q1 20 delivery contracts appear overvalued, given a forecast end-October carryout near 3.8 tcf (a near 600 bcf y/y surplus). A cold event of considerable duration and deviation from normal will be necessary to substantially move the needle on the Feb-19 and Mar-19 contracts, but the window of time for this to occur is closing, especially with market expectations coalescing near a 1.5 tcf carryout. Today’s (10 January) trading, nearly flat d/d, does not appear to have taken stock of either a larger-than-expected withdrawal or any potential cold in the forecast.
Timing risk to Mexican pipeline projects (on which we have taken a conservative view on start-up) and LNG projects (start-up timing of which we align with FERC granting permission and the typical amount of time to next steps, such as fuel gas to feedgas) continue to throw a potential bearish pall over balances, possibly delaying demand from both types of exports.
Outside of delays to Mexican pipeline and LNG exports, two components of our balances could be absorptive of supply—lower-than-expected Lower 48 production and anaemic net trade with Canada. Some initial 2019 guidance has been released on the back of weakness in both the oil and gas complexes. These producers include Chesapeake, Parsley, Antero and Diamondback. For producers with a Permian bias, the effect on gas output may not be nearly as noticeable as on oil, given that acute gas evacuation limitations at play had already led to subdued gains in our forecast. However, for gas production tied to infrastructure demand charges on new pipes, gas output may prove stickier.
Antero’s 2019 guidance came in 0.1 bcf/d or so below street estimates. However, growth is still up by some 20% y/y, and the producer has commitments of 0.7 bcf/d on Mountaineer Xpress (MXP) at demand charges of $0.40/mmbtu for flows from Sherwood onto TCO Pool, and $0.60/mmbtu for Sherwood flows onto Leach XPress.
Columbia Gas Transmission is still awaiting FERC authorisation to start the MXP pipe, although the initial request filed on 12 December 2018 would put just the 0.7 bcf/d for Antero into service. For other Appalachia pipelines—including Rover, NEXUS and Atlantic Sunrise in 2018—FERC waited until approximately 70% of environmental restoration in an area was complete before allowing any infrastructure to come online. Spreads 3-8 of MXP—which Columbia Gas Transmission has filed to start up—had an average of 71% of such remediation complete by the end of December, a figure that is only set to increase as more work is completed. In a report filed with FERC on 31 December 2018, Columbia Gas Transmission noted it planned to complete restoration on those spreads by mid-January before finishing work on the other 57 miles of pipe (which have not yet been fully lowered into place). As such, there is a strong likelihood that MXP will see its initial 0.7 bcf/d come into service before end-January.
We are currently guiding for a reduction in net Canadian trade of 0.4 bcf/d y/y. We anticipate WCSB production will be down by 0.2 bcf/d y/y during injection season 2019 based on drilling and permitting activity (see Canada Monthly: Arrested production, 28 November 2018) and TransCanada’s preliminary maintenance schedule. The schedule appears to be somewhat lighter y/y, targeting 140 days vs 175 last year. Additionally, it indicates that the events will not be as impactful as in 2018.
However, given the schedule is preliminary and TransCanada’s track record of extensive maintenance in previous years, we believe there is more downside than upside risk to this figure. Our balances assume a reversion to 10-year normal weather and include the fundamental backdrop of US inventories not in need of incremental supply. These fundamentals, together with the potential for more maintenance than initial plans indicate—especially with the current lack of a preliminary schedule from the Westcoast Pipeline system—point to downside risk to net trade.
Puts are dominating an increasingly bearish option market, as per CME and NASDAQ block trades. Earlier this week, we observed 3,000 lots of Apr-19 $1.75/mmbtu put traded, likely one of the first ‘1-handle’ trades of the injection season, underscoring that potential bearish bias. While we are not calling winter off just yet, each passing week is currently keeping our end-March inventories afloat and thereby eliminating scarcity concern. At this point, our pillar of balances is transitioning away from a view on how low March inventories may be, to a path to a very robust end-October carryout.
|Fig 1: Projected EIA weekly storage change, bcf||Fig 2: Projected storage y/y change, bcf/d|
|Source: EIA, Energy Aspects||Source: EIA, Energy Aspects|